Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact firstname.lastname@example.org. THE MIOCENE TREES OF THE ROCKY MOUNTAINS PROFESSOR T. D. A. COCKERELL University of Colorado. The living arborescent flora of the Rocky Mountain region is at the present time occupying the attention of a number of able workers, including Nelson in Wyoming, Rydberg of the New York Botanical Garden, Sudworth of the Forest Service, Ramaley, Bethel and Schneider in Colorado, Wooton in New Mexico, and others. As a re- sult of all this activity, we are promised two manuals of Rocky Mountain botany, and a third of trees alone, so we shall have three separate and independent treatments of our woody flora to compare and choose from. Unfortunately, those who have been so active and ex- haustive in their investigations of the living flora have not cared, as a rule, to consider the historical or paleo- botanical side of the subject. Many " recent " botanists seem to have a positive dislike for fossil plants, and few manifest any great eagerness to receive information about the ancestors or predecessors of the species which occupy their attention. Like all enthusiasts, the writer is filled with the idea that the matter has only to be ade- quately presented to command universal attention; and hence offers this discussion, not so much for the paleo- botanists as for those students of living plants whose active interest may be aroused in the problems involved. Going back from the present time, we are practically without information concerning the plants of our region until we come to the Florissant beds, assigned to the Miocene. These beds, however, contain an abundant series of remains, many of the plants beautifully pre- served, as the accompanying illustrations show. They testify to a climate both warmer and damper than that of the present day, the arborescent genera including Sapindus, Ficus, 1 Diospyros, Persea, Leucasna, Anona, 1 The determination of Ficus is based on the leaves. In confirmation of *> i 32 THE AMERICAN NATURALIST [Vol. XLIV etc., but so far as known no palms. Some, as Ailantlins americcma, pertain to genera now restricted to Asia. The determination of the age of the Florissant beds has been a matter of some difficulty, notwithstanding the large number of organisms preserved. Comparing the flora with that of the European Tertiary, I have felt satisfied that it should be referred to the Miocene, and probably to the Upper Miocene. The resemblance to the flora of GEningen in Baden, known to be upper Miocene, is most striking. Thus we have the following parallel or representative species : Florissant. (Eningen. Liquklambar eonvexum Ckll. Liquiclambar europccum A. Br. TJlmus braiinii Heer, Lx. Ulmiis braunii Heer. Comptonia insignis (Lx.) Ckll. Comptonia amingensis A. Br. For ana spcirii Lx. Porana cening crisis A. Br. Parana tenuis Lx. Porana macrantlia Heer. Acer florissanti Kirch. Acer tri cuspid latum A. Br. 2 Many others could be cited. On the other hand, the Florissant incense cedar, Heyderia or Libocedrus colo- radensis Ckll., is to be compared with H. salicornioides, of the Low T er Miocene of Radoboj in Croatia. The Florissant redwood, Sequoia haydeni (Lx.), is not related to 8. sternbergi Heer from GEningen, but to 8. langsdorfii (Brgt.) Heer of the Swiss Lower Miocene; this species, however, survived into the Upper Miocene in Italy and Galicia. This 8. langsdorfii has been recognized in America also from the Upper Cretaceous to the Miocene, and some of the Florissant specimens have been referred to it; but the identity of the plants from so many diverse localities and horizons is questionable, and from Floris- sant I think we have only one species, 8. haydeni. The Sequoia and Libocedrus of Florissant are both very closely related to their living Californian allies; so it comes a discovery by Mr. Brues, who in working over the parasitic Hymen- optera from Florissant has come upon what appears to be a genuine fig- insect,, apparently of the South American genus Tetrapus Mayr. -Acer trilobatum (Sternb., 1825) A. Br., 1845; not A. trilobatum Lam., 1786. No. 517] MIOCENE TREES 33 much so that one is in some difficulty to point out any tangible differences. This is equally true of a number of other cases, of which the following are illustrative : Florissant. Living. Finns ivlieeleri Ckll. Finns flexilis James. Finns sPurgisi Ckll. Finns tee da L. Ailanthus americana Ckll. Ailanthus glandnlosa L. Sambucus newtoni Ckll. Sambucus arboreseens Nutt. Anona spoliata Ckll. Anona glabra L. Bobinia brittoni Ckll. Bobinia pseudacacia L. Fopuhis lesquereuxi Ckll. Fopulus angust-i folia James. Quercus lyratiformis Ckll. Quercus 1 grata Walt. Sapindus coloradensis Ckll. Saj)indus drummondi H. & A. So numerous are the resemblances to the living flora that one might well feel persuaded to refer the beds to the Pliocene— certainly better there than to the Oligocene or Eocene ! However, the Florissant fishes, with the exception of Anna, are of extinct genera, and no less than 178 genera of insects are supposed to be extinct. For a variety of reasons, based chiefly upon a study of the insects, I believe that the Florissant period corre- sponds with Oshorn's "Fifth Faunal Phase" (Bull. 361, II. S. G-eol. Survey), in which a new fauna was invading the country from Eurasia, while connection with South America had not yet been established. Some of the Florissant groups of insects, such as the Aphididre and Bombyliida?., seem to represent the original American fauna uncontaminated ; while others show old world types, the most significant and interesting of which is the tsetse fly (Glosshia)* Osborn's "Fifth Phase " in- cludes the Middle and Upper Miocene, and so far as may be judged, Florissant should belong near the middle of it. The attempt to correlate the Florissant beds with other American floras ascribed to the Miocene brought out a number of difficulties. With the exception of the little- 3 A second species of tsetse fly, Gloss ha osborni Ckll., lias been recently discovered. It is only ICU mm. long, the wing 7 mm.; the venation is normal for the genus, but the first basal cell bulges less subapically than in Scud- der\s species. 34 THE AMEBIC AN NATURALIST [Vol. XLIV Fig. 2. Weinmannia lesquereuxi Ckll. Fig. 1. Weinmannia phenacophylla Ckll. known formation at Elko Station, Nevada, I do not find anything which really seems to correspond with Floris- sant. According to the theory outlined above the Mascall beds of Oregon, which possess a varied flora, should be either contemporaneous or (more probably) somewhat earlier. Fortunately, fourteen species of mammals have been obtained from the Mascall, and these place it rather definitely in the Middle Miocene. Considering, therefore, a probable moderate difference in time, combined with noteworthy geographical and altitudinal differences, we ought to find the Mascall flora similar to, but by no means identical with, that of Florissant ; and this is exactly what comparisons show. Thus of the 77 Mascall plants (nearly all trees) re- ferred to definite genera, no less than 56 are congeneric with those of Florissant. Of those not congeneric, five are so dubious that they have not been specifically deter- mined. The Mascall genera not yet found at Florissant are the following: 1. Equisetum.— This has no significance, as it abounds in Colorado to-day, and must have been present during the Florissant period. No. 517] MIOCENE TREES 35 2. Ginkgo.— Represented in the Mascall by a fragment not specifically determined. This genus is not known in the Rocky Mountains later than the Laramie and Liv- ingston, on the border line between the Cretaceous and Tertiary. As is well known, there is a single living (Asiatic) species. 3. Thuites.—A fragment not specifically determined. It is practically identical with T. elireuswardi Heer (Miocene of Sachalin and Spitsbergen), but that plant appears to be referable to the modern genus Chamrecy- paris. 4. Glyptostrob'iis.—A genus still living in China. It was supposed to occur at Florissant, but I believe the material so referred all belongs to Sequoia. The Mascall material is not above suspicion of also being Sequoia; indeed Lesquereux so referred one of the specimens. 5. Taxodiwn. — The Mascall specimens are referred by Knowlton to the widely distributed T. disticlmm mio- cenum Heer, which should be called Tax odium disticlmm dubium = Taxodinm dubium (Sternb.) Heer, originally described from Bilin. This differs from Sequoia by the deciduous leaves, which are not decurrent at the base as in Glyptostrobus. The genus still lives in our southern states. 6. Artocarpiis.—'ReipTesented by very fragmentary material, doubtfully referred to A. californica Kn. 7. Magnolia.— Major Bendire collected a plant which Knowlton says "may well be" M. ingle fi-eldi Heer. It has not been obtained by recent collectors. Magnolia dayana Ckll. ined. (M. lanceolata Lx. 1878, not Link. 1831) is listed by Knowlton as from the Mascall, but in his detailed account he says it is from Cherry Creek, which should be Lower Eocene. 8. Lawns.— Florissant has a species of Persea ; Lauras and Persea are allied, and not distinctly separated by paleobotanists. 9. Platanus.—The Mascall specimens appear to belong 36 THE AMERICAN NATURALIST [Vol. XLIV Si^ ffih t 3? ' JESS ^r^kmmmm ' A Jkk jlffr \kSfc Mir ^L m^ pi ' jjr ^* . nt • v^ *d &** Fig. 3. Sequoia hayileni (Lesquereux). Redwood. to three species, but none are sufficiently well preserved for positive specific identification. 10. Pnmus.— The two Mascall species described by Knowlton are only doubtfully referred to this genus, which is of course abundant in the modern flora. 11. Rulac— Generic reference rather uncertain; the genus is scarcely separable from Acer, which occurs at Florissant. 12. 2E sculus.— This well-known living genus is repre- No. 517] MIOCENE TREES 37 sented in the Mascall by leaflets which closely resemble an nnclescribed Florissant species which may be a Ber- beris, but is certainly not an iE senilis. 13. (jren'ia. — The Mascall plant is referred by Knowl- ton to U. crenata (linger) Heer, which occurs in Europe at (Eningen. 4 Three other genera, Phragmites, Oyperacit.es and Smilax, are non-arborescent, and have no particular sig- nificance. Tims it would appear that in the Middle Miocene period Ginkgo and Glyptostrobns— if we may accept the determinations— had not yet retreated from the Amer- ican continent, but survived at least in the northwest. For the rest, the Mascall flora is no doubt a lowland one as compared with that of Florissant, and this alone would explain many of the differences ; thus, no one would ex- pect to find Taxodinm growing around a mountain lake. Dr. Knowlton has described (Monog. II. S. Geol. Sur- vey, Vol. 32, part 2) an extensive flora from the Yellow- stone, which he regards as Miocene. The fossil plants of the Yellowstone National Park are divided by him into three series: (1) Fort Union, which is Basal Eocene, (2) Intermediate, said to be Miocene, and (3) Lamar Flora, also Miocene. With the first we are not now con- cerned, but the others must be compared with the flora of Florissant. Considering the relative proximity of the Yellowstone beds to those of Colorado, one would expect to find much similarity and even identity in the plants ; but this is not the case. The difference of locality, with a moderate difference in time, might perhaps account for the diversity of species; but the Yellowstone flora as a whole does not impress one as being so modern as that of the Mascall beds or Florissant, while there is a sig- nificant identity of species with those of the Eocene. I have extracted from Knowlton 7 s tables a list of all the Yellowstone "Miocene" plants said to occur else- where or in the Eocene, with the following result: 4 The African Grewia crenata Hoclist., 18(38 (not linger, 1850), takes tlie name G. po-pulifolia Vahl, 1790. 38 THE AMERICAN NATURALIST [Vol. XLIV Fig. 4. Rhus coriarioides Lesquereux. Sumach. 1. Common to Fort Union (Eocene), Intermediate and Lamar. Sequoia langsdorfii (Brgt.). Said to go down to the Laramie (Cretaceous). Juglans rugosa Lx. Goes down to the Laramie. Castanea pulchella Kn. Ficus densifolia Kn. Lmirus calif omica Lx. Also auriferous gravels of California. Lmirus grandis Lx. (not Wallich). Also aurif- erous gravels of California. Platanns guillelmcd Gopp. Perhaps also Laramie. Aralia notata Lx. Also Denver beds. * Elceodendron polymorphtim Ward. 2. Common to Fort Union and Intermediate. Equisetum canalicidatum Kn. Perhaps also in the Lamar. No. 517] MIOCENE TREES 39 Magnolia (?) pollardi Kn. * Ulmtis minima Ward? Sapindus affinis Newby. 3. Common to Fort Union and Lamar. Asplenium iddingsi Kn. Lyg odium Jcaidfussi Heer„ Eq-uisetum decidmim Kn. Juglans crescentia Kn. Fiats asimincefolia Lx. Also auriferous gravels of California. Laurus prUnigenia linger? Malapoenna lamarensis Kn. Sapindus grcmdifoliohis "Ward. Sapindus warclii Kn. * Ilicoria antiqiia (Newt).). *Ulmus pseudofidva Lx.? Those marked with an asterisk occur in the Fort Union only outside of the Yellowstone. 4. Common to the Intermediate and the Denver beds (Basal Eocene). Osmunda affinis Lx. 5. Common to the Lamar, Basal Eocene and Laramie. Rhammis rectinervis Heer, Lx. Heer describes this from Monod, in the Lower Miocene ; we may venture to doubt the identity of the American plant. Thus we have twenty-six plants specifically identical with those of the Basal Eocene. 5 6. Common to Lamar and "Green River" of Knowlton. (See also under 7.) Salix elongata 0. Web. Said to occur at Elko Station, Nevada, but represented only by un- characteristic fragments. The determination of 5 The Mascall is supposed to have five species common to the Fort Union; but of these two are doubtful, two others are the conifers Sequoia Jangs- clorfii and Tax odium, while the fifth is Sapindus outusifolius, to which a single specimen from the Mascall "seems to belong." S. obtusifolius was originally described from beds supposed to belong to the Washakie (Later Eocene). 40 THE AMERICAN NATURALIST [Vol. XLIV Fig. 5. Ulmus hilliw Lesquereux. Elm. the Lamar plant is considered doubtful by Knowlton. Fagus (Fagopsis) longifolia (Lx.). Elko Station, Nevada; Florissant (very abundant) and Eocene (?) of British Columbia. The British Columbia locality is on the Similkameen River, whence come various fossil insects. Dr. Daw- son (quoted by Scudder) considered these de- posits Miocene. The Yellowstone collection in- cludes about forty specimens which Knowlton No. 517] MIOCENE TREES 41 s refers here, all from Fossil Forest Ridge. Thi is, undoubtedly, a distinctively Miocene plant, and must be accepted as pertinent evidence. The determination must be presumed to be correct, though it may be pointed out that various other leaves have almost exactly the same venation and appearance. This is especially true of the species of Zelkova, to which genus Engler (1894) actually referred F. longifolia, though the discovery of the fruit has since shown that it is not related thereto. Vlmus plw'inervia, as figured by Heer from Alaska, is also almost exactly like F. longifolia; it is considered doubtfully Eocene, but Knowl- ton has recognized it in the Mascall (Miocene). From the shape of the base, and other features, it seems to me certain that the Alaskan plant is not the original U. plurincrvia, of which linger gives four figures in the Chloris Protogsea. The latter is decidedly more elm-like in appearance. Con/his macquarrii (Forbes) Heer. This plant, as recognized in America, is a Fort Union and possibly Laramie species ; recorded also from the Eocene (?) of Alaska. Diospyros hracliysepala A. Br. As recognized in this country, this is a Laramie and Fort Union species; the record from Florissant I believe to be erroneous. None of the above belong to the genuine Green River series ; three are quite without significance as indicating Miocene affinities, but the Fag us stands out as a solitary Miocene representative. 7. Common to the Lamar and the Auriferous gravels of California. (See also under 1 and 3.) Juglans leonis Ckll. Two specimens in the Lamar. Popuhis balsamoides Gopp. Also Miocene (?) of Alaska. Known in the Yellowstone only from a fragment, which certainly can not be positively determined as balsamoides: in fact, it shows 42 THE AMERICAN NATURALIST [Vol. XLIV Fig. 6. Myrica drymeja (Lesquereux). some differences, at least as compared with the original European balsamoides, which ought to be specific. Salix varians Gopp. Eocene (?) of Alaska. The Lamar plant is a fragment, and according to the figure, the margin is quite unlike that of the European varians. Salix angusta A. B. Said to occur also in. the Basal Eocene and true Green River. The Lamar material consists of doubtful fragments. Quercus furcinervis americana Kn. Ficus shastensis Lx. ? Ficus sordida Lx. A mere fragment from the Lamar. Ficus asimincefolia Lx. Very indifferent material from the Lamar. Also Fort Union. Magnolia calif ornica Lx.? The Lamar plant is represented by a single specimen, "so much No. 517] MIOCENE TBEES 43 broken that its positive identification is not pos- sible" (Knowlton). Per sea pseudocarolinensis Lx. The Lamar speci- men figured, "the best one found," consists of the upper half of a leaf ; what there is of it ap- pears to agree with the Californian species, al- though it has more lateral viens. Rhus mixta Lx.? Aralia ivhitneyi Lx. Also in the Intermediate. None of the Yellowstone specimens are perfect, but they appear to belong to this handsome species. Thus the species common to the Lamar and Auriferous gravels, but not known from Basal Eocene, are few, and in several cases of doubtful identity. As the reference of the Lamar to the Miocene rests wholly on the resem- blance of the flora to that of the Auriferous gravels, with the exception of the indication afforded by Fagus longi- folia, it must be considered at least somewhat dubious. It is also to be remarked that eleven species of plants are supposed to be common to the Yellowstone Fort Union and the Auriferous gravels, although two of these, at least, are doubtfully from the gravels, while in four or five cases the Yellowstone material is fragmentary or doubtful. It is one thing, however, to recognize distinct elements in common between the Auriferous gravels and the Lamar, and another to prove the latter Miocene thereby. The former may be conceded, the latter I think not. Lesquereux enumerates thirteen species from the Au- riferous gravels which are almost identical with living species ; he also cites seventeen which are evidently, but not very closely, related to living ones. Of the thirteen, four are enumerated from the -Lamar; of the seventeen, not one. Of the four common to the Lamar, three are dubious, and only Juglans leonis (a species represented to-day by the Asiatic J. regia) appears to be of satis- factory standing. 44 THE AMERICAN NATURALIST [Vol. XLIV Fig. 7. Populus crassa (Lesquereux). Cottonwood; probably fruit of P. lesqucrcuxi. Fig. 8. Populus lesquereuxi Ckll. Cottonwood. Four species of the Auriferous gravels are said by Lesquereux to be identical with Miocene plants, but are all unsatisfactory, as follows: (1) Fag us antipofii; per- haps goes to the Laramie, and the Californian specimen was only half a leaf. (2) Populus zaddachi; supposed to go down to the Basal Eocene. (3) Ficus tilicefolia;" Fig. 0. Salix ramaleyi Ckll. Willow. c Ficus tilicefolia (A. Br.) Heer, 1856, has priority over F. tilicefolia Baker, Jn. Linn. Soc. 21: 443 (1885), from Madagascar. The latter may become Ficus oakeriana n. n. No. 517] MIOCENE TREES 45 said to go down to the Laramie. (4) Aralia zaddachi; of uncertain determination, one of the specimens was Platanus dissecta. None of these is found in the Lamar, hut F.antipofii is in the Yellowstone Fort Union. Bight other species from the Auriferous gravels are stated to he allied to Miocene species, five of these being also related to living plants. One of the five, Juglans oregoniana, has since proA T ed to he from the Mascall, and not to occur in the Auriferous gravels. The other three are as follows : Fic-us sordida Lx. Allied to, or perhaps identical with, F. grmnlandica of Greenland. A frag- ment referred to this has been found in the Lamar. Ficus menscc n. n. (F. micropliylla Lx., 1878, not Salzm., Mart. Fl. Braz. 4: 93). Allied to F. planicostata '—but this is a species of the Basal Eocene and Laramie. Aralia wliitneg-i Lx., said to be allied to an Evans- ton species, which would be Eocene. It is thus apparent that the Auriferous gravels flora has no decisive Miocene affinities, but is composed of two sets of plants, one related to living forms, the other to those of the Eocene. It is known to be a mixed lot, and when I recently suggested to Dr. J. C. Merriam, of the "University of California, that it might perhaps be partly Pliocene and partly Eocene, he replied that this might indeed be the case. It is further to be remarked that Knowlton formerly regarded the Mascall flora as having affinity with that of the Auriferous gravels; but he subsequently discovered that certain of the species he had most relied on were really confined to the Mascall, and did not occur in the gravels at all. "This correlation therefore fails," he states, and the absence of relationship stands as an argu- ment against the Miocene age of the gravels. The conclusion seems to be legitimate that the Yellow- stone Intermediate and Lamar florae are Upper Eocene, or at least older than Miocene. Were they really Mio- 46 THE AMERICAN NATURALIST [Vol. XLIV Fig. 10. Ptelea modcsta (Lesquereux). Fig. 11. Melia expulsa Ckll. cene, with so much resemblance to even the Basal Eocene, the Florissant flora, to get as far on the other side as its lack of affinity would suggest, would have to be projected somewhere into the future! If this opinion is in any degree correct, Florissant remains as the only Rocky Mountain locality for Miocene plants, so far as known. The accompanying figures, all taken from specimens obtained at Florissant by the University of Colorado ex- peditions, will give a good idea of the material from that locality. Nowhere else in America are Tertiary plants so well preserved. As compared with the Eocene flora, and especially the Basal Eocene, the Florissant trees are more diverse in type, with usually smaller leaves, which are often compound. Excessively moist condi- tions are not indicated, though there was evidently much more moisture than at the present day. Some of the plants are even somewhat xerophytic, indicating that the higher slopes may have been relatively dry. Osborn remarks on the evidence of increasing summer droughts No. 517] MIOCENE TREES 47 in the Middle Miocene. So far as the mammals are con- cerned, this is chiefly indicated by the plains fauna. Ow- ing to the generally higher temperature, the air was probably moister than at present, but the moisture may have carried farther, to be precipitated on the mountains. Thus the conditions on the plains and towards the sea may have resembled those of Southern and Lower Cali- fornia to-day, with a comparatively damp atmosphere but little or no preciptation during a considerable part of the year. The desert fauna and flora of the southwest is a highly specialized one, which has certainly not come into existence since the Miocene, at least as regards its fundamental types; so it becomes necessary to postulate a desert region during Miocene times, and no doubt much earlier. Whether we shall ever know much about the Tertiary deserts from fossil remains is perhaps question- able, though we certainly have evidence of a semi-desert fauna, as is illustrated by the large tortoises of the Upper Miocene. The Florissant beds afford us a wonderful insight into the mountain life of the Miocene, and must have a continually increasing significance in relation to the evolution of the fauna and flora of this continent. Most unfortunately, they have as yet yielded no recogniz- able mammalian remains, but I am convinced that these will eventually be found. The beds are far from being exhausted, and comparatively little digging has been done at the place where fragments of a mammal were obtained— a locality which I shall be glad to describe in detail to any one who cares to go and try his luck. In the meanwhile, large collections both of plants and of in- sects, already obtained, remain to be investigated and re- ported upon, but for various reasons the work proceeds slowly.